Download chapter4

Chapter 4 : Transducers - Generation and Detection of Ultrasound
I.
Introduction

Transducers are used as both transmitters and receivers, converting electrical
energy to acoustical energy and vice versa. In low frequency applications (below
20KHz), microphones and loudspeakers are two well-known examples. For
diagnostic ultrasound, higher frequencies are required and piezoelectric materials
are most commonly used.

Piezoelectricity is defined as the generation of an electrical polarization in a
substance by the application of a mechanical stress and, conversely, a change in
the shape of a substance when an electric field is applied. In other words, a
material is strained when an electric field is applied to it.

Commonly used piezoelectric materials include naturally occurring crystals, such
as quartz and certain man-made ceramic materials, such as lead
zirconate-titanates (PZTs). Crystals such as quartz are inherently piezoelectric,
with properties determined by their crystallographic features. In contrast,
man-made ceramics are polarized above the Curie temperature (typically around
320-370oC for PZTs) by the application of strong electric fields to induce
anisotropy responsible for their strong piezoelectric properties. PZTs are the most
commonly used piezoelectric materials for diagnostic ultrasonic imaging.

In addition to PZTs, PVDFs and composites are two other commonly seen
materials in medical ultrasound. PVDF is often used for acoustic field
measurements due to its broad bandwidth and sufficiently small thickness. Note
that small thickness is necessary in order to minimize the interference of sound
field due to the presence of the hydrophone. Composite materials, on the other
hand, have gained wide interest due to the potential of improving imaging
performance.
II. Piezoelectric Constitutive Relations

Consider the following figure, in which the equilibrium spacings between
neighboring rows of atoms are L, a1 and a2, and q is the magnitude of the charge
of the atoms,
Chapter 4
25
P
P+P
a2
a2+a2
a1
a1+a1
no strain
L
L
with strain
the polarization (per unit volume) of the object is
P

dipole strength of unit cell q ( a2  a1 )
.
 2
volume of unit cell
L ( a2  a1 )
In the presence of strain S, a1 and a2 change to a1+a1 and a2+a2, respectively,
and the polarization changes to P+P. Since a1=a1S and a2=a2S, under the
conditions that a1 and a2 are small, we obtain
P  PS  eS ,
where e is defined as the piezoelectric stress constant. The total change in
electric displacement (or electric flux density) in the presence of an electric field
E is
D  E  P  E  eS ,
where  is the permittivity with zero or constant strain. We find that the electric
displacement in a piezoelectric material is dependent on both the electric field and
the strain.

Similar to the above derivation, we can determine the stress in a piezoelectric
medium due to an electric field E. Since the forces per unit area on the positive
and negative atoms are qE/L2, the stresses in the regions of length a1 and a2 are
therefore
Chapter 4
26
qE
L2
.
qE
T2  2
L
T1  
The average internal stress in the medium due to the electric field is
TE 
a1T1  a2T 2
 eE .
a1  a2
The total stress ( T piezo ) applied to the object is the sum of the externally applied
stress T (or T nonpiezo ) and the internal stress TE (or T elect ). According to Hooke’s
law and defining c E as the elastic constant (under a constant electric field
condition), we have
T  TE  c E S
T  c E S  eE
(or T piezo  T nonpiezo  Telect ).

Equations D  E  eS and T  c E S  eE are known as piezoelectric
constitutive relations. The above derivation assumed one-dimensional situations,
where a single scalar e is adequate to represent the coupling between the elastic
and electric properties.
III. Wave Propagation in Piezoelectric Materials

Equations governing the acoustic wave propagation in a piezoelectric material are
obtained by using Newton’s second law with the constitutive relations. Based on
previous derivation of one-dimensional wave propagation in a non-piezoelectric
material, we can replace the bulk modulus B by the elastic constant c E and
 2 w( z, t )  eE ( z, t ) / c E )  
T ( z, t ) / c E
 2 w( z, t )

(
c
/

)

(
c
/

)


.
E
E
2
z
z
t 2
 z

Re-arranging the above equation, we obtain
Chapter 4
27
 2 w( z, t )   2 w( z, t ) 1  eE ( z, t ) 
.


cE
cE
z
z 2
t 2

Imagine the piezoelectric material in the form of a plate with metal electrodes on
each face. If the electrodes on opposite faces are short circuited, the electric field
is reduced to zero and the above equation is reduced to a homogeneous wave
equation. On the other hand, if the electrodes are open-circuited or if the medium
is infinite long in z, there would be no free charges in the transducer medium (i.e.,
D is a constant in z, but may be a function of time). Under this condition, the
above equation takes the following form by applying the other constitutive
relation
 2w ( z , t )   2w ( z , t )
e D ( z , t ) e 2  2w ( z , t )
e 2  2w ( z , t )





B
c E
z
c E
c E
z 2
t 2
z 2
z 2

 2w ( z , t ) 

2
2

z
 B 1  e / c E


nonpiezo
Note that cvelocity


.
  2w ( z , t )

0
2


t

B
represents the propagation velocity in a

non-piezoelectric medium. From the above equation, we have
piezo
velocity
c
c
nonpiezo
velocity

e2 
1 

 c E 
1/ 2
.
It is then obvious that acoustic waves propagate at a higher velocity when the
material is being piezoelectrically stiffened. The elastic constant under a constant
electric displacement condition ( c D ) is related to that under a constant electric
field condition ( c E ) by

e2
c D  c E 1 
 c E

 .

e2
The quantity k , where k 
, is known as the electromechanical coupling
c E
2
constant. The magnitude of the electromechanical coupling constant is a useful
index of the strength of the piezoelectric effect in a particular material. Note that
 is the dielectric permittivity under a constant strain condition.
Chapter 4
28
IV. Piezoelectric Generation and Detection of Ultrasound

Ultrasonic waves are generated by the application of an external electric field to a
piezoelectric material. A previously derived inhomogeneous wave equation
indicates that a gradient in eE (the product of the piezoelectric stress constant and
the electric field) is the source for the generation of mechanical disturbances.
Usually the surfaces of a piezoelectric material offer the sharpest discontinuity in
both e and E, hence they are the strongest sources of sound. This can be shown by
the following simplified figure. Assuming a parallel plate capacitor containing a
piezoelectric material of dielectric constant .

0
D
E
L
0
(a)

L
(b)
0
L
(c)
Consider the application of a charge density to the capacitor shown in figure (a).
Since there are no free charge between the plates, we have
D
 0 and hence D
z
is a constant as illustrated in the figure (b). Additionally, the permittivity is
typically higher inside the piezoelectric material, thus making the electric field E
smaller than E in the outside. Therefore, figure (c) can be used to represent E and
it becomes apparent that the only gradient in the electrical field occurs at the
surfaces of the material. Consequently, the surfaces of the piezoelectric material
are the predominant sources for the generation of ultrasound.

Piezoelectric detection of ultrasonic waves is reciprocal to the process of wave
generation. In other words, the conversion of mechanical energy into electrical
energy is also a phenomenon dominated by the behavior at the surfaces of the
piezoelectric material. As shown below, the voltage measured across a
piezoelectric plate is the integral of the electric field over the thickness of the
crystal
L
V ( t )   E ( z , t ) dz
0
Chapter 4
29
V(t)
impinging
wave
piezoelectric
medium
electrode
surface area A
electrode
z=0
z=L
Using the constitutive relation relating the electric field to the strain and the
electric displacement, the voltage becomes
V ( t )  
L
0
L D ( z ,t )
e
S ( z , t ) dz  
dz
0


From Gauss’s law, we have D ( z , t ) 
q (t )
where q ( t ) is the total charge on
A
area A . Therefore, the above equation reduces to
e L w ( z , t )
q (t )
dz 

0

z
A / L
.
e
q (t )
  w ( L , t )  w ( 0 , t ) 

C0
V (t )  
where C 0 
A
is the capacitance of the piezoelectric plate. The magnitude of
L
the net displacement can be calculated from the boundary conditions placed at the
surfaces of the material. In addition, the above equation shows that under open
circuit situations, i.e., q ( t ) is not time-varying, the voltage developed across
the plate is directly related to the relative displacements of the front and back
surfaces of the material. If the thickness of the plate corresponds to an odd
integral number of half-wavelengths of the ultrasonic wave impinging on the
material (i.e., the two surfaces oscillate 180o out of phase), then the relative
displacement of the front and back surfaces, w ( L , t )  w ( 0 , t ) , is the largest. In
contrast, if the thickness of the crystal is an even number of half-wavelengths,
then the amplitude of oscillation of the two surfaces is in phase, and therefore,
w ( L , t ) w ( 0 , t )  0 .
V. Equivalent Circuits
Chapter 4
30

Based on the piezoelectric detection equation that we previously derived, we have
e
q (t )
.
V ( t )   w ( L , t ) w ( o , t ) 

C0
Taking the partial derivative with respect to time on both sides, we obtain
i (t )  C 0
V ( t )
e
 C 0 u ( L , t )  u ( 0 , t )
t

which can be represented by the following equivalent circuit.
i(t)
+
C0
V(t)

C0e/(u(L,t)-u(0,t))
Defining force as we do voltage in electrical circuits, and particle velocity as we
do current in electrical circuits, the non-piezoelectric component can be described
e q (t )
using the other constitutive relation (i.e., T ( z , t ) piezo  T ( z , t ) nonpiezo 
)
 A
and represented by the following circuit (a transmission line)
I2
I1
+
V1
Z1
Z2
+
V2
Z3
where Z 1  Z 2 due to symmetry. Let Z 0 represent the characteristic
impedance of the piezoelectric material, we can obtain the following equations
using methods similar to those used by deriving the acoustic wave equations
V ( z ,  )  V1 (  ) e  jz / c V2 (  ) e jz / c
I ( z , ) 

1
Z0
V (  ) e
1
 jz / c
V2 (  ) e jz / c

.
In order to obtain Z 1 and Z 3 , the following relations need to be used.
Chapter 4
31
Z3 
V2
I1
I 2 0

V
Z1  Z3  1
I1
V ( L , )
I ( 0 , )
I ( L ,  )0
.
I 2 0
Letting I 2  0 , the following equation yields
Z 0 I ( L ,  )  0  V1 (  ) e  jL / c V2 (  ) e jL / c .
Therefore
V2 (  )  V1 (  ) e 2jL / c .
and
Z3  Z0

V1 (  ) e  jL / c  e  jL / c

V1 (  ) 1  e
 2jL / c


Z0
  jZ 0 cos ec2L / 
j sin 2L / 
Furthermore, Z 1 (= Z 2 ) can be obtained as the following:
Z1 

V ( 0 , )
I ( 0 , )
I 2 0
 Z3  Z0
cos 2L /   1
 jZ 0 tan L / 
j sin 2L / 
By using a transformer to couple the electrical components with the acoustic
components, we then obtain the following equivalent circuit (a.k.a. Mason
Equivalent Circuit)
u(0,t)
u(L,t)
Z1
Z1
Z3
-C0
1:N
i(t)
Chapter 4
32
C0
N=C0e/

With the front and back faces of the transducer being loaded by mechanical
impedances Z F and Z B , the equivalent circuit can be re-drawn as the following:
-C0
Z3
C0

Z1
Z1
ZB
ZF
1:N
The above circuit can be further transformed to the following circuit identity with
Z4=1/Z1.
-C0
2Z4
C0

2ZB
2Z1
-ZB
2ZB
ZF
1:2N
In the neighborhood of a mechanical resonance, impedance of the branch
containing 2Z1 is large and the branch can be neglected. Therefore, the model can
be approximated with the following circuit. Such an approximation is adequate to
describe the operation of a transducer near resonance driving a wide variety of
mechanical loads.
-C0
2Z4
ZB
Chapter 4
C0
1:2N
33
ZF

An alternative equivalent circuit (KLM Equivalent Circuit) has been developed to
be better suited for broadband operations. Please refer to the following paper if
interested.
- D. Leedom, R. Krimholtz and G. Matthael, “Equivalent circuits for transducers
having even or odd symmetry piezoelectric excitation”, IEEE Trans. on Sonics
and Ultrasonics, vol. SU-18, No. 3, pp. 128-141, July 1971.
VI. Design Considerations for Broadband Transducers

Broadband transducers are necessary for pulse-echo imaging applications in order
to achieve high range resolution, which is inversely proportional to the pulse
bandwidth. However, a short pulse is usually achieved by sacrificing sensitivity.
Considering the following piezoelectric transducer with no matching or damping
layers, it rings and produces an unacceptable long pulse.
body
PZT
By placing a lossy material (highly attenuating) which has a similar acoustic
impedance as the PZT, the reflection at the back of the transducer can be reduced
and therefore the pulse can be shortened. Apparently, sensitivity is degraded due
to attenuation.
body
PZT
backing
The acoustic impedance of a typical PZT material is around twenty times higher
than that in the body, therefore, part of the sensitivity loss can be recovered by
adding one or multiple quarter wave matching layers at the front surface of the
transducer.
Chapter 4
34
body
PZT
backing
matching layer

Two-way insertion loss, defined as the ratio of the available electrical power
generated by the device as a receiver to the electrical power dissipated in the
device as a transmitter under the conditions in which the acoustic wave produced
is reflected from a perfectly reflecting interface and received by the same
transducer, is often used as a measure of the electromechanical efficiency of the
transducer. The lower the insertion loss is, the higher the sensitivity can be
achieved.

Equivalent circuits can be used to match both the mechanical impedance and the
electrical impedance. Ideally, real part of the electrical impedance should be
50Ohms in order to match a typical transmitter output impedance. Imaginary
part of the impedance, on the other hand, should be to zero in order to obtain
optimal efficiency. This is typically done by placing an inductor (as a tuning
element) to cancel the capacitance of the transducer (and sometimes the
transducer cable). However, the improvement in sensitivity is often gained at the
price of bandwidth of the spectrum.

An acoustic lens is often placed on the front of the transducer in order to provide
a fixed geometric focusing. This is particularly important for imaging using
one-dimensional arrays, in which case the geometric focusing is provided along
the non-scan direction.

A transducer array consists of many piezoelectric elements. The elements are
arranged depending on specific scan formats. A typical diagram of a
one-dimensional transducer array is shown in the following.
Chapter 4
35
Chapter 4
36